The invention discloses an intelligent parameter
optimization system and method for mine
millisecond blasting, and relates to the technical field of mine
millisecond blasting datamation modeling and intelligent optimization control. The method is used for solving the problems that quantitative tracing of blasting parameters is difficult, the effect is difficult to predict, and multi-target collaborative optimization and closed-loop
verification are difficult under hard constraints such as vibration or flying rocks. The method comprises the following steps: based on multi-
source data such as single convergence drilling, charging,
detonation,
vibration measurement, image and surveying and mapping of blasting operation records, constructing a sample set with unified fields and units; training and calibrating the blasting effect prediction model to output the PPV, the lumpiness and the farthest distance between the lumpiness and the flying stone; taking cost, vibration, lumpiness and flying rocks as multiple targets, introducing a threshold value and enforceable constraints, solving a recommendation scheme, and generating a disturbance instance based on an offset combination form
library to perform conservative quantile recheck; and finally, determining
executable millisecond blasting parameters through field
verification and feedback updating.